Multidimensional Patterns Research Articles

Novel Multidimensional Feature Pattern Classification Method and Its Application to Fault Diagnosis

With the development of modern process industries, data-driven fault diagnosis methods have attracted more and more attention. In this paper, a novel nonlinear fault diagnosis method based on multidimensional feature pattern classification (MDFPC) is proposed. The proposed MDFPC method integrates multikernel independent component analysis (MKICA) with adaptive rank-order morphological filter (AROMF). First, some dominant independent components capturing nonlinearity are extracted from the historical process data using the MKICA algorithm, getting the template signal and the testing signal of each fault pattern. Then, a multidimensional signal classification method based on AROMF is developed to achieve the diagnosis of fault patterns. The effectiveness of the proposed fault diagnosis method is demonstrated by carrying out a case study using the Tennessee Eastman process.

Influence of personal meaning organization and 5-HTTLPR genotype on cortisol stress reactivity in healthy women

IntroductionReactivity to acute psychosocial stress in the framework of a physiological multidimensional pattern affects several individual-level systems that include genetic factors and features related to personality development. The 5-HTTLPR genotype has been implicated in the modulation of susceptibility to environmental stimuli.ObjectivesIn the present study, 91 healthy young women were investigated (i) for their reactivity to a standardized psychosocial laboratory stressor (TSST), as measured by changes in salivary cortisol; (ii) in terms of 5-httlpr genotype and (iii) in terms of their personality profile according to the post-rationalist personal meaning organizations (PMOs), which are considered as adaptive modes of response to environmental stressors.MethodsParticipants were divided into three 5-HTTLPR genotype groups (s/s; s/l, and l/s). The quantitative and qualitative variables that may affect circulating cortisol were compared among the three groups. A multiple linear quantile regression analysis was then performed to evaluate the effect of the personality profile, as Outward/Inward PMO, and 5-HTTLPR genotype on the median level of cortisol, considered as dependent variable.ResultsComparison of the variables that may affect circulating cortisol no significant differences. Salivary cortisol changed significantly in the course of the TSST. Reactivity to stress was affected by personality profile and the 5-HTTLPR genotype and also by body mass index and age.ConclusionsThe present data suggest that the psychosocial stress response is a multidimensional physiological event that is affected by a variety of factors as diverse as 5-HTTLPR genotype, personality profile, BMI, and age.Disclosure of interestThe authors have not supplied their declaration of competing interest.

L’émotion–douleur : de l’intérêt de considérer la douleur comme une émotion

ObjectivesPain is a common experience, both physical and emotional. However we often feel powerless with our patients suffering pain. This paper aims to give a new heuristic and psychological understanding of pain. MethodologyAccording to new theories, recent researches as well as different points of view, we form an analogy between pain and emotion. Throughout historical considerations pain has always been perceived through theories and beliefs, changing its definition. This is also the case for emotion. Could they be two ways of expressing a single phenomenon? ResultsFirst, we must clarify the definition of emotion. In past, emotion was considered as a multiple-conditioned notion. To be considered as an emotion the pain had to fill numerous features, which differ according to the scientific opinions. The emotion may be considered as a physical expression or perceived only as the consequences of a real emotion, i.e., the subjective feeling. We propose as a way of thinking that emotion brings together these two concepts. We support a flexible vision of emotion. To investigate the field of the emotion different mental steps may be thought of: we should conceive of the emotion as a stimulus, as an emotional evaluation and as a tendency to action, which becomes an emotional response. These steps are colored by subjective feelings. It can be summarized in three levels: the situation decoding (1), the response organization (2) and the effectiveness of the response (3). Second pain can be considered as a complex notion involving personal and subjective feelings. We can use multidimensional patterns and consider emotion with its multiple features: the generating mechanisms, the pain perception, the pain behavior and the environment. Each stage can be divided in different ways. Hence pain treatment could be approached as an emotional treatment. Indeed, we can make a link between generating mechanisms and emotion situation decoding, between pain perception and emotion situation decoding and response organization, between pain behavior and environment and emotion response effectiveness and consequences. A heuristic analogy can be formed enriching therapeutic possibilities. First, emotion and pain could be considered as a unique phenomenon that can be expressed in different ways. We can let aside the opposition between psychological and physiological: these are two different levels of a same phenomenon. This point of view can be helpful to treat pain disorders because the skills to manage emotion disorders are well known.

Profiles of adolescent religiousness using latent profile analysis: Implications for psychopathology.

Prior research has documented robust associations between adolescent religiousness/spirituality (R/S) and psychopathology outcomes including externalizing and internalizing symptomatology, yet no previous studies have examined these associations with adolescent R/S profiles using a person-centred approach. We examined whether there are identifiable subgroups characterized by unique multidimensional patterns of R/S experiences and how these experiences may be related to externalizing and internalizing symptomatology. The sample consisted of 220 Appalachian adolescents between 12 and 18years old who were primarily White and primarily Christian. Latent profile analysis revealed three profiles of adolescent R/S: high religiousness (28.4%), introjectors (47.6%), and low religiousness (24.0%). These profiles were differentially related to internalizing and externalizing symptomatology such that the high religiousness group was significantly lower than the introjectors with respect to internalizing and externalizing symptomatology and lower than the low religiousness group in externalizing symptomatology. Implications and suggestions for future research using person-centred approaches to better understand differential developmental trajectories of religious development are provided. Statement of contribution What is already known Prior research has demonstrated a negative relationship between adolescent religiousness and spirituality (R/S) and psychopathology. Numerous studies document the differential relationships between aspects of R/S and psychopathology; however, few have done so from a person-centred perspective. There are several theories that outline how R/S to study R/S when paying specific attention to culture. Saroglou's Big Four dimensions of religion (believing, bonding, behaving, and belonging) posits that these four dimensions (1) are able to delimit religion from proximal constructs; (2) translate major distinct dimensions of religiousness; (3) can be seen across cultural contexts; and (4) are good candidates to study cultural variability in religion due to their diversity; however, to the authors' knowledge there has been no attempt to synthesize the Big Four dimensions and person-centred work. What the present study adds The present study found three profiles of adolescent R/S: high religiousness, low religiousness, and of particular interest, the introjectors. Those high in introjection seem to have a partial internalization of religiousness due to their low score in private practices but moderate to high scores on other aspects of religiousness. This group would not have been found through the use of traditional data analysis techniques or even through structural equationmodels. Importantly, those in the introjector group were also significantly higher in internalizing symptomatology than those in the high religiousness group, and higher in externalizing symptomatology than both the high religiousness and low religiousness. This 'u-shaped' pattern in which those in the middle-range of R/S were the worst off would also not have been found using traditional data analysis techniques.

Scale-up of nature\u2019s tissue weaving algorithms to engineer advanced functional materials

We are literally the stuff from which our tissue fabrics and their fibers are woven and spun. The arrangement of collagen, elastin and other structural proteins in space and time embodies our tissues and organs with amazing resilience and multifunctional smart properties. For example, the periosteum, a soft tissue sleeve that envelops all nonarticular bony surfaces of the body, comprises an inherently “smart” material that gives hard bones added strength under high impact loads. Yet a paucity of scalable bottom-up approaches stymies the harnessing of smart tissues’ biological, mechanical and organizational detail to create advanced functional materials. Here, a novel approach is established to scale up the multidimensional fiber patterns of natural soft tissue weaves for rapid prototyping of advanced functional materials. First second harmonic generation and two-photon excitation microscopy is used to map the microscopic three-dimensional (3D) alignment, composition and distribution of the collagen and elastin fibers of periosteum, the soft tissue sheath bounding all nonarticular bone surfaces in our bodies. Then, using engineering rendering software to scale up this natural tissue fabric, as well as multidimensional weaving algorithms, macroscopic tissue prototypes are created using a computer-controlled jacquard loom. The capacity to prototype scaled up architectures of natural fabrics provides a new avenue to create advanced functional materials.

The application of a topological clustering technique to capture forms and dynamics of deprivation in contemporary Switzerland

This paper proposes the application of an innovative technique of clustering and projection—the Self-Organizing Map—in order to identify multidimensional patterns of deprivation in contemporary Switzerland. Starting from a rich set of non-monetary indicators from the Swiss Household Panel, a topological map has been extrapolated. This map is composed of 32 micro-clusters that are subsequently grouped into four macro-clusters, each of which expresses a specific profile. The analysis carried out indicates that about one out of two individuals living in Switzerland enjoy a state of multidimensional well-being, one in three shows signs of slight vulnerability, one in five is in a condition of consistent vulnerability and one in twenty is deprived. After having outlined the space of multidimensional patterns, the movements of these individuals have been reconstructed from 1999 to 2012 and the role of “true path dependence’ have been measured.

Multidimensional mutual ordering of patterns using a set of pre-trained artificial neural networks

The article shows that large artificial neural networks can be used for mutual ordering of a set of multi-dimensional patterns of the same nature (handwritten text, voice, smells, taste). Each neural network must be pre-trained to recognize one of the patterns. As a measure of ordering one can use the entropy of patterns “Strangers” that are input to a neural network trained to recognize only examples of the pattern “familiar”. The neural network after training reduces the entropy of the examples of the pattern “Familiar” and increases the entropy of examples of pattern “Stranger.” It is shown that the entropy measure of the ordering always has two global minima. The first minimum corresponds to the pattern “Familiar”, the second to the inversion of the pattern “Familiar”. It is also shown that the Hamming distance between the patterns belonging to two different groups (groups of the two global minima) is always as large as possible.

Fractionation of muscle activity in rapid responses to startling cues.

Movements in response to acoustically startling cues have shorter reaction times than those following less intense sounds; this is known as the StartReact effect. The neural underpinnings for StartReact are unclear. One possibility is that startling cues preferentially invoke the reticulospinal tract to convey motor commands to spinal motoneurons. Reticulospinal outputs are highly divergent, controlling large groups of muscles in synergistic patterns. By contrast the dominant pathway in primate voluntary movement is the corticospinal tract, which can access small groups of muscles selectively. We therefore hypothesized that StartReact responses would be less fractionated than standard voluntary reactions. Electromyogram recordings were made from 15 muscles in 10 healthy human subjects as they carried out 32 varied movements with the right forelimb in response to startling and nonstartling auditory cues. Movements were chosen to elicit a wide range of muscle activations. Multidimensional muscle activity patterns were calculated at delays from 0 to 100 ms after the onset of muscle activity and subjected to principal component analysis to assess fractionation. In all cases, a similar proportion of the total variance could be explained by a reduced number of principal components for the startling and the nonstartling cue. Muscle activity patterns for a given task were very similar in response to startling and nonstartling cues. This suggests that movements produced in the StartReact paradigm rely on similar contributions from different descending pathways as those following voluntary responses to nonstartling cues.NEW & NOTEWORTHY We demonstrate that the ability to activate muscles selectively is preserved during the very rapid reactions produced following a startling cue. This suggests that the contributions from different descending pathways are comparable between these rapid reactions and more typical voluntary movements.

Disentangling the health benefits of walking from increased exposure to falls in older people using remote gait monitoring and multi-dimensional analysis

Falls and physical deconditioning are two major health problems for older people. Recent advances in remote physiological monitoring provide new opportunities to investigate why walking exercise, with its many health benefits, can both increase and decrease fall rates in older people. In this paper we combine remote wearable device monitoring of daily gait with non-linear multi-dimensional pattern recognition analysis; to disentangle the complex associations between walking, health and fall rates.One week of activities of daily living (ADL) were recorded with a wearable device in 96 independent living older people prior to completing 6 months of exergaming interventions. Using the wearable device data; the quantity, intensity, variability and distribution of daily walking patterns were assessed. At baseline, clinical assessments of health, falls, sensorimotor and physiological fall risks were completed. At 6 months, fall rates, sensorimotor and physiological fall risks were re-assessed. A non-linear multi-dimensional analysis was conducted to identify risk-groups according to their daily walking patterns.Four distinct risk-groups were identified: The Impaired (93% fallers), Restrained (8% fallers), Active (50% fallers) and Athletic (4% fallers). Walking was strongly associated with multiple health benefits and protective of falls for the top performing Athletic risk-group. However, in the middle of the spectrum, the Active risk-group, who were more active, younger and healthier were 6.25 times more likely to be fallers than their Restrained counterparts.Remote monitoring of daily walking patterns may provide a new way to distinguish Impaired people at risk of falling because of frailty from Active people at risk of falling from greater exposure to situations were falls could occur, but further validation is required. Wearable device risk-profiling could help in developing more personalised interventions for older people seeking the health benefits of walking without increasing their risk of falls.

Time course and neural basis of early visual perception for facial emotions

The complex multidimensional visual patterns of facial expression is particularly well suited for studying perception and the processes underlying perceptual integration of component parts into a meaningful whole. With the high resolution in milliseconds, event-related potential (ERP) may be an excellent technique to track the time course of cognitive processing for facial emotions. Facial affective stimuli elicit particular ERP components involved in early processing of facial expressions, including: early sensory processing of visual stimuli: the P100; early structural encoding of facial features: the N170; second-order sustained activities necessary for consequent deeper processing: the posterior P2 and decoding of facial features to identify emotional content: the N250. We reviewed above mentioned ERP components respectively, in order to add the knowledge about ERP components as indicators of neural dynamics of facial emotion processing, and to provide a point of reference for further exploration on facial emotion processing that may be useful for understanding human cognition development and pathology. Key words: Face; Emotion; ERP; P100; N170; P2; N250

Analyzing the Social Knowledge Construction and Online Searching Behavior of High School Learners During a Collaborative Problem Solving Learning Activity: a Multi-Dimensional Behavioral Pattern Analysis

The present study aims to explore the behavioral patterns of the social knowledge construction process and the online searching behaviors in a collaborative problem solving learning activity for high school students, and further compares the different behavioral patterns of the high- and low-performing teams. A total of 78 high school students from two schools participated. This study applied sequential analysis to analyze the students’ social knowledge construction and online searching behavior from a large amount of screen-recording data. Interestingly, the results indicate that social interaction that is irrelevant to the discussion task is significantly correlated with academic-related discussion content. Reaching a higher cognitive level of social knowledge construction (e.g., reaching agreement or applying newly constructed meaning) contributes to a successful team project. For online searching behavior, the high-performing teams exhibited systematic online searching behavior and concentrated on the task, while the low-performing teams displayed chaotic searching behavior and were distracted from the task at hand, and seemed to rarely propose their searched results or ideas in their discussions. Based on the results, several possible explanations and suggestions are proposed including the need to promote more adaptive motivation and to provide scaffolding for collaborative skills.

Multidimensional patterns of metabolic response in abiotic stress-induced growth of Arabidopsis thaliana.

Contextualization of specific transcriptional responses of Arabidopsis within the stress-tissue-time perspective provides a simplified representation of the cellular transcriptional response pathways to abiotic stress, while reducing the dimensions in gene-oriented response description. Crops resistant to abiotic stresses are a long-term goal of many research programs, thus understanding the progression of stress responses is of great interest. We reanalyzed the AtGenExpress transcription dataset to go beyond gene-level characterization, and to contextualize the discrete information into (1) a process-level signature of stress-specific, time-specific, and tissue-specific responses and (2) identify patterns of response progression across a time axis. To gain a functional perspective, ∼1000 pathways associated with the differentially-expressed genes were characterized across all experiments. We find that the global response of pathways to stress is multi-dimensional and does not obviously cluster according to stress, time or tissue. The early response to abiotic stress typically involves induction of genes involved in transcription, hormone synthesis and signaling modules; a later response typically involves metabolism of amino acids and secondary metabolites. By linking specific primary and secondary response pathways, we outline possible stress-associated routes of response progression. The contextualization of specific processes within stress-tissue-time perspective provides a simplified representation of cellular response while reducing the dimensions in gene-oriented response description. Such simplified representation allows finding stress-specific markers based on process-combinations pointing whether a stress-specific response was invoked as well as provide a reference point for the conductance of comparative inter-plant study of stress response, bypassing the need in detailed orthologous mapping.

Subject-specific strength and weaknesses of fourth-grade students in Europe: a comparative latent profile analysis of multidimensional proficiency patterns based on PIRLS/TIMSS combined 2011

In 2011 the Progress in International Reading Literacy Study (PIRLS) and the Trends in International Mathematics and Science Study (TIMSS) were conducted at fourth grade in a number of participating countries with a shared representative sample. In this article we investigate whether there are multidimensional proficiency patterns across the competency domains or not. In order to derive proficiency patterns across the reading (PIRLS), mathematics and science (TIMSS) competence domains, latent profile analyses (LPA) of students’ plausible values were conducted. For this, the grade four student sample from 17 countries were combined and analyzed. The international reference model that resulted from this analysis was then applied with constraints to all 17 countries separately so that substantial comparisons between countries became possible. To describe and compare the differences between national profiles a classification system was developed and applied to all countries’ profile patterns. As a result of these international LPA seven groups of learners were identified. The profiles were approximately equidistant and parallel. For all countries we find that achievement across domains can be explained by a general level of achievement rather than subject-specific strengths or weaknesses of learners. However, subject-specific strengths and weaknesses can be identified but are—with the exception of Malta and Northern Ireland—for most of the countries rather small. For only about half of the countries, a rather uniform pattern of subject-specific strengths and weaknesses can be found on all competence levels. The subject itself varies between countries. In the other countries high, intermediate and low achievers differ in their relative subject-specific strength and weaknesses. The results suggest that differences in average achievement in TIMSS and PIRLS should also on country level be interpreted with caution. International comparative studies should further investigate potential reasons for the differences between countries.

Self-organization and autonomy: Emergence of degrees of freedom in dynamical systems

Approached from the point of view of the basic processes that constitute the self-organization of living systems, autonomy means the generation of identity and the minimal unity of a system, as a consequence of the self-production of internal components and processes of an organism, self-regulation of its internal variables, and self-sustaining of its internal resources. However, a living system is also a dynamical system, which means that the emergence of identity and the unity of the system is inseparable from the generation of its degrees of freedom. These degrees of freedom have different levels of complexity, given by the multidimensional patterns instantiating them, offering various alternatives to respond to environmental perturbation. From the point of view of the multidimensionality of degrees of freedom of a living system, which depends on the degree of self-organization and complexity of the organism, one can distinguish three types of autonomy: minimal or basic autonomy, sensorimotor autonomy, and strong autonomy. Put in these terms, autonomy depends on the abilities of the organism to access some degrees of freedom of higher complexity, to enhance its degrees of freedom by its coupling with the environment, as a result of its bodily skills, and to consciously control and monitorize its degrees of freedom, as a result of its higher-order cognitive abilities. Keywords: self-organization, autonomy, degrees of freedom, dynamical system, autopoietic system.

Arrhythmia classification using Mahalanobis distance based improved Fuzzy C-Means clustering for mobile health monitoring systems

In this paper, an improved electrocardiogram (ECG) beats classification system is proposed, which is based on Fuzzy C-Means (FCM) clustering algorithm. The classification of ECG beats is necessary in order to diagnose the type of arrhythmia (e.g., Atrial Premature Contraction (APC), Premature Ventricular Contraction (PVC), Right Bundle Branch Block (RBBB) etc.) present in the ECG records. The efficiency of any classification model highly depends on the “most relevant” set of features used. The primary goal of this study is to classify different arrhythmic beats with reduced set of relevant-only ECG attributes. The attribute selection model is based on Mahalanobis-Taguchi System (MTS); a multi-dimensional pattern recognition tool, which can dynamically choose the important set of ECG features. The number of most relevant features can vary from person to person according to the type of arrhythmia present in the respective ECG signals. A traditional Euclidian Distance (ED) based FCM can detect the spherical clusters but it may lead to improper clustering in some cases. As a solution to this problem, Mahalanobis Distance (MD) is used in the proposed model in order to improve the distance measurement procedure. In our proposed system, MD based improved Fuzzy C-Means (FCM-M) clustering is used to classify the arrhythmic beats. Experimental results show that the performance of FCM-M is significantly better than the conventional FCM for arrhythmia classification. Another direction of our proposed research is to use the concept of initial cluster centroid in order to reduce the number of program iterations. In our experiments, the number of program iterations is reduced to an average of 53% when initial centroid is assigned to FCM-M with the same classification results.

Mapping frequent spatio-temporal wind profile patterns using multi-dimensional sequential pattern mining

ABSTRACTHolistic understanding of wind behaviour over space, time and height is essential for harvesting wind energy application. This study presents a novel approach for mapping frequent wind profile patterns using multi-dimensional sequential pattern mining (MDSPM). This study is illustrated with a time series of 24 years of European Centre for Medium-Range Weather Forecasts European Reanalysis-Interim gridded (0.125° × 0.125°) wind data for the Netherlands every 6 h and at six height levels. The wind data were first transformed into two spatio-temporal sequence databases (for speed and direction, respectively). Then, the Linear time Closed Itemset Miner Sequence algorithm was used to extract the multi-dimensional sequential patterns, which were then visualized using a 3D wind rose, a circular histogram and a geographical map. These patterns were further analysed to determine their wind shear coefficients and turbulence intensities as well as their spatial overlap with current areas with wind turbines. Our analysis identified four frequent wind profile patterns. One of them highly suitable to harvest wind energy at a height of 128 m and 68.97% of the geographical area covered by this pattern already contains wind turbines. This study shows that the proposed approach is capable of efficiently extracting meaningful patterns from complex spatio-temporal datasets.

Multidimensional Profiling of Task Stress States for Human Factors: A Brief Review.

This article advocates multidimensional assessment of task stress in human factors and reviews the use of the Dundee Stress State Questionnaire (DSSQ) for evaluation of systems and operators. Contemporary stress research has progressed from an exclusive focus on environmental stressors to transactional perspectives on the stress process. Performance impacts of stress reflect the operator's dynamic attempts to understand and cope with task demands. Multidimensional stress assessments are necessary to gauge the different forms of system-operator interaction. This review discusses the theoretical and practical use of the DSSQ in evaluating multidimensional patterns of stress response. It presents psychometric evidence for the multidimensional perspective and illustrative profiles of subjective state response to task stressors and environments. Evidence is also presented on stress state correlations with related variables, including personality, stress process measures, psychophysiological response, and objective task performance. Evidence supports the validity of the DSSQ as a task stress measure. Studies of various simulated environments show that different tasks elicit different profiles of stress state response. Operator characteristics such as resilience predict individual differences in state response to stressors. Structural equation modeling may be used to understand performance impacts of stress states. Multidimensional assessment affords insight into the stress process in a variety of human factors contexts. Integrating subjective and psychophysiological assessment is a priority for future research. Stress state measurement contributes to evaluating system design, countermeasures to stress and fatigue, and performance vulnerabilities. It may also support personnel selection and diagnostic monitoring of operators.

Actively heated high‐resolution fiber‐optic‐distributed temperature sensing to quantify streambed flow dynamics in zones of strong groundwater upwelling

AbstractZones of strong groundwater upwelling to streams enhance thermal stability and moderate thermal extremes, which is particularly important to aquatic ecosystems in a warming climate. Passive thermal tracer methods used to quantify vertical upwelling rates rely on downward conduction of surface temperature signals. However, moderate to high groundwater flux rates (>−1.5 m d−1) restrict downward propagation of diurnal temperature signals, and therefore the applicability of several passive thermal methods. Active streambed heating from within high‐resolution fiber‐optic temperature sensors (A‐HRTS) has the potential to define multidimensional fluid‐flux patterns below the extinction depth of surface thermal signals, allowing better quantification and separation of local and regional groundwater discharge. To demonstrate this concept, nine A‐HRTS were emplaced vertically into the streambed in a grid with ∼0.40 m lateral spacing at a stream with strong upward vertical flux in Mashpee, Massachusetts, USA. Long‐term (8–9 h) heating events were performed to confirm the dominance of vertical flow to the 0.6 m depth, well below the extinction of ambient diurnal signals. To quantify vertical flux, short‐term heating events (28 min) were performed at each A‐HRTS, and heat‐pulse decay over vertical profiles was numerically modeled in radial two dimension (2‐D) using SUTRA. Modeled flux values are similar to those obtained with seepage meters, Darcy methods, and analytical modeling of shallow diurnal signals. We also observed repeatable differential heating patterns along the length of vertically oriented sensors that may indicate sediment layering and hyporheic exchange superimposed on regional groundwater discharge.

GLOBAL LAND COVER CLASSIFICATION USING MODIS SURFACE REFLECTANCE PROSUCTS

The objective of this study is to develop high accuracy land cover classification algorithm for Global scale by using multi-temporal MODIS land reflectance products. In this study, time-domain co-occurrence matrix was introduced as a classification feature which provides time-series signature of land covers. Further, the non-parametric minimum distance classifier was introduced for timedomain co-occurrence matrix, which performs multi-dimensional pattern matching for time-domain co-occurrence matrices of a classification target pixel and each classification classes. The global land cover classification experiments have been conducted by applying the proposed classification method using 46 multi-temporal(in one year) SR(Surface Reflectance) and NBAR(Nadir BRDF-Adjusted Reflectance) products, respectively. IGBP 17 land cover categories were used in our classification experiments. As the results, SR and NBAR products showed similar classification accuracy of 99%.

GLOBAL LAND COVER CLASSIFICATION USING MODIS SURFACE REFLECTANCE PROSUCTS

Abstract. The objective of this study is to develop high accuracy land cover classification algorithm for Global scale by using multi-temporal MODIS land reflectance products. In this study, time-domain co-occurrence matrix was introduced as a classification feature which provides time-series signature of land covers. Further, the non-parametric minimum distance classifier was introduced for timedomain co-occurrence matrix, which performs multi-dimensional pattern matching for time-domain co-occurrence matrices of a classification target pixel and each classification classes. The global land cover classification experiments have been conducted by applying the proposed classification method using 46 multi-temporal(in one year) SR(Surface Reflectance) and NBAR(Nadir BRDF-Adjusted Reflectance) products, respectively. IGBP 17 land cover categories were used in our classification experiments. As the results, SR and NBAR products showed similar classification accuracy of 99%.